3 Global social justice The possibility of social justice beyond states in a world of overlapping practices Ayelet Banai, Miriam Ronzoni, and Christian Schemmel

Introduction The claim that broadly egalitarian principles of social justice1 apply globally has been the focus of a far-reaching debate, recently described as "one of the most intense controversies in political philosophy" (Van Parijs 2007: 638). Positions range from the view that principles of social justice apply exclusively within states to the altogether opposite contention that rejecting the global application of domestic principles of social justice means rejecting the fundamental moral axiom that all human beings have equal worth. Thomas Nagel, for example, opposes the global application of principles of egalitarian social justice, arguing that such principles are valid only within the framework of coercive state institu-tions of which we are putative joint authors (Nagel 2005). On the other hand, some versions of cosmopolitanism claim that any model short of global equality of some sort constitutes unjustifiable discrimination among individuals on the basis of morally arbitrary factors such as geographical location or citizenship (e.g. Caney 2001b).

In this chapter, we defend a particular approach to this debate, namely the practice-dependent method - which takes existing social and institutional practices to be constitutive of principles of justice - and explore some of its implications for social justice beyond state borders. We begin by locating the methodological problem within the global justice debate. We then explore the advantages of the practice-dependent approach over its practice-independent rivals. Finally, we argue that the practice-dependent approach, properly understood, yields a nuanced intermediate position between statism and global egalitarianism.

Arguing that standards of egalitarian social justice do not apply globally does not, as such, imply holding that no principles of justice exist among and beyond states. Theorists who reject the global application of principles of egalitarian social justice tend to agree that some other principles of justice do hold. For instance, David Miller raises substantial objections to global egalit-arianism (Miller 2005), but defends the "idea of a global minimum that is due to every human being as a matter of justice" (Miller 2007: 231). Humanitarian

Global social justice 47 aid on the occasion of natural disaster relief, for example, becomes on this account a duty of justice. In Michael Blake's (2001) argument egalitarian principles only apply within states, in order to justify the restriction of auton-omy that arises from states' comprehensive coercive power. At the same time, according to this argument, the value of autonomy also yields implications for the global context, namely that justice requires minimal subsistence for every-one. More generally still, liberal statists agree upon principles of just interna-tional interaction, including non-aggression, compliance with treaties, avoidance of harm to non-nationals as well as specified duties of assistance (Butt 2009: 65-72).

The above picture of within-state and between-state justice brings into view an additional aspect of the debate: why are states so special? Confining a theory of justice to formal state institutions seems problematic, when in reality people significantly impact one another across borders through a whole variety of other social, economic, cultural and political structures and organizations (O'Neill 2000; Buchanan 2000). According to Charles Bertz's and Thomas Pogge's seminal argument, the level of interconnectedness within the global order is such that we share a worldwide basic structure which, among other things, deeply and systematically affects people's life prospects and the fulfilment of human rights, through, among other things, regulating the world economy and global trade. The importance of this position, contested though it may be, is that it provides an institutionalist account of the responsibility of individuals to all other individuals worldwide, and an institution-based justification for the claim that (at least some) principles of social justice should apply globally (Beitz 1979; Pogge 2008a, 1994). Moreover, a moderately cosmopolitan institutionalist account such as the one suggested by (the later) Pogge need not lead to the conclusion that domestic and global basic structures are alike and give rise to the same normative demands (Cohen and Sabel 2006; Ronzoni 2008 and 2009).

This debate is both essential and insightful because the disagreements in question have emerged among philosophers who tend to agree on fundamental liberal egalitarian commitments concerning domestic justice. Indeed, at the outset, much of this debate consisted in attempts to work out the implications of John Rawls' A Theory of Justice for the global scenario. The fact that Rawls clarified his own position, in his The Law of Peoples, and took a stand against global social justice, is by no means the last word. Rawls has left enough space to rethink his own conclusions, as his arguments rely upon the highly contested factual hypothesis that states are generally self-sufficient and discrete entities.

Furthermore, the participants in the global justice debate do not only agree on a liberal egalitarian framework of domestic justice, but they also seem to share some basic convictions about the nature of the global problem. Nagel, whilst advocating a state-confined application of principles of egalitarian justice, notes that:

The gruesome facts of inequality in the world economy are familiar. Roughly 20 percent of the world's population lives on less than a dollar a

48 A. Banai et al. day, and more than 45 percent live on less than two dollars a day, whereas the 15 percent who live in the high-income economies have an average per capita income of seventy-five dollars a day.

(Nagel 2005: 118) Pogge cites similar facts about striking poverty and glaring inequalities in support of the argument for the global scope of social justice (Pogge 2001c: 328). The sharp disagreements that have emerged in spite of those common starting points encourage us to dig deeper in order to grasp why such conflicting conclusions have been reached. Two fundamental questions, it emerges, are at the bottom of most of the disagreements. First is the problem of redefining the modern legal state. If states are not the self-sufficient, sovereign, territorial legal units that they were often thought to be, what are they then? Do they retain a particular and morally relevant role? Second, methodological assumptions on how normative theories of justice ought to be developed have come to the fore and are currently being reassessed. This chapter focuses on the second question, and proposes to approach the first one through it.

This core methodological question concerns the role of existing social and political practices in identifying normative principles of justice. More specifi-cally, since - as mentioned above - much of the debate about the global applica-tion of principles of social justice has evolved around Rawls' theory of justice, the problem lies in identifying the role of practices in Rawls' approach. Although never without contestation, many scholars assumed for a long time that Rawls arrived at his two principles for an ideal theory of justice under practice-independent assumptions. Rawls' question was largely seen as being the follow-ing: assuming that individuals consider one another free and equal (regardless of existing practices), to which principles would they agree behind a veil of igno-rance?2 However, as Aaron James puts it, "Rawls is clear that ideal theory requires starting from 'the international political world as we see it '" (James 2005b: 281-2). Rawls, James argues, is committed to a practice-dependent account of justice: principles of justice are principles that aim to govern specific existing practices, and their content therefore depends on the nature of said practices.

In the remainder of this chapter, we unpack and defend the methodological assumptions of the practice-dependent approach to justice. Moreover, we argue, in contrast to the so far prevailing position among advocates of a practice-dependent methodology, that a commitment to this approach does not imply an endorsement of the view that problems of social justice only arise within the borders of the state. To the contrary, the practice-dependent approach can helpfully highlight that there are, indeed, problems of global social justice - namely, problems that concern the fundamental terms of global rule-governed social interaction and cooperation beyond the establishment of a global minimum and just interaction between states. Such problems and the normative principles they call for, however, are not identical to their domestic counterparts.

Global social justice 49 Practice-dependence vs. practice-independence The practice-dependent approach (henceforth PDA) has been put forward in a series of articles over the last few years (e.g. Sangiovanni 2008, 2007; Ronzoni 2009; James 2005b; Valentini forthcoming).3 The main aim of these contribu-tions has been to sketch the contours of PDA as a distinct approach to the identi-fication of principles or conceptions of justice. Thus far, proponents of PDA have not engaged in exploring its advantages over its competitor - the practice-independent approach (henceforth PIA). This section takes on this task, outlining reasons to prefer PDA to PIA for the purpose of developing principles of social and global justice.4 The main thrust of the argument is that such reasons are methodological rather than substantively moral. PDA takes salient facts about the global order - namely those facts that characterize its basic practices - into account from the outset, at the level of developing fundamental principles of justice. PIA, instead, relegates such facts to the level of application of fundamen-tal principles. We shall argue that the latter strategy leaves us not only without reasonable orientation as to how that application ought to take place, but, and more importantly, without a yardstick for assessing the plausibility of those very principles in the first place.5

PDA, as said above, takes its cue from Rawls. It does not, however, begin with arguments from the original position, but focuses on the social practices to which arguments from the original position are supposed to apply. Famously, in the case of A Theory of Justice and Political Liberalism, these social practices are described as the "basic structure" of a domestic society - that is, its main social, political, and economic institutions taken together as a unit. In Rawls' theory, this framework is not itself chosen in the original position; it is presup-posed. In general, Rawls defines as a "practice", or "social institution", "any form of activity specified by a system of rules which define offices, roles, moves, penalties, defenses, and so on, and which gives the activity its structure." Exam-ples are "games and rituals, trials and parliaments, markets and systems of prop-erty" (Rawls 1999c: 47, no. 1).

Hence, the fundamental idea of PDA is that conceptions of justice have to start from an account of the practices they are supposed to apply to.6 Sangio-vanni defines this as the "practice-dependence Thesis: The content, scope, and justification of a conception of justice depends on the structure and form of the practices that the conception is intended to govern" (Sangiovanni 2008: 138). The task of the political philosopher is hence not simply to come up with uncon-strained moral arguments against which any existing social practices have to be measured. It is rather a twofold task, consisting of first identifying and interpret-ing the nature of such practices; and then, once this has been achieved, of deliv-ering moral arguments for the best conception of justice that ought to govern these very practices, given their nature and purpose. James and Sangiovanni explain this task by drawing on Ronald Dworkin's constructive interpretation (Dworkin 1986, ch. 2; Sangiovanni 2008: 148ff.; James 2005b: 301ff.), accord-ing to which normative theory has to proceed in three steps. First, a practice is

50 A. Banai et al. identified as the one to which both interpretation and normative arguments are to apply, such as the basic structure of domestic societies, or the regime of interna-tional law. Second, we should deliver an interpretation of the normative purpose of that practice, from the point of view of those participating in it. In the case of the basic structure, Rawls argues that its point is to set and enforce the terms of cooperation that make the production of primary goods - namely those goods which everyone needs in order to carry out her conception of the good -possible. The third step, finally, seeks to answer the question which principles of justice would best regulate the pursuit of the purpose of the specific practice in question, and how the practice ought to be reformed, extended, or restricted on this basis. Views based on PDA aim to be consistent with the basic liberal prin-ciple of the equal moral worth of all human beings (Sangiovanni 2007: 3). According to PDA, however, abstract principles such as this do not qualify as principles of justice proper, but merely constitute general moral requirements that every conception of justice has to respect. The task of a suitable conception of justice is then to provide an account of what it means, for a specific practice, to respect such requirements.

As the above explanation sought to clarify, PDA is a methodology for singling out principles of justice; it is not itself a moral position, such as statism, which claims that requirements of social and distributive equality are only relevant in the context of the institutions of the state. This distinction is crucial, but has largely been overlooked. The current debate about global justice has often failed to keep these two levels apart, because PDA has mostly been used in arguments for the claim that the scope of principles of egalitarian social justice is restricted to the state. This conclusion follows from PDA only on the basis of additional empirical claims, namely, that practices which make demanding requirements of social justice both possible and necessary exist only at the state level (Sangio-vanni 2007: 6). As several critics have pointed out, such empirical claims are con-troversial (Abizadeh 2007; Barry and Valentini 2009; Caney in this volume). PDA, we suggest below, may indeed lead to the conclusion that principles of social justice apply beyond states, even if different ones from those that apply within them. The need to provide a more adequate account of the relevant prac-tices does not, however, undermine the appropriateness of the methodology.

But why bother with PDA in the first place? Global egalitarianism, after all, proceeds on the basis of a much simpler methodology: global egalitarian princi-ples of justice can be derived from the principle of the equal moral worth of all individuals, coupled with the claim that it is morally arbitrary into which state one was born (Caney 2001b, 2005a; Beitz 1983; Tan 2004). Moreover, the argu-ment for PIA goes, focusing on existing social practices stands in the way of identifying injustice and is biased towards the status quo, because it overlooks the need for justification for having practices like states in the first place, with their massive effects on global inequality.

It is important to note that a commitment to PIA does not necessarily imply that one has to regard significant inequalities between individuals as always unjust, regardless of whether any relationship or interaction exists among them.

Global social justice 51 Relational views - such as the one according to which domestic egalitarian prin-ciples apply globally as soon as agents (individuals or otherwise) interact with each other across borders in ways that significantly affect their interests, or have pervasive impact on each other in some other way - also fall under PIA (e.g. Abizadeh 2007). Such views still differ from PDA for two reasons. First, they rely on relatively unspecified ideas of interdependence and pervasiveness, rather than on accounts of specific social practices that are governed by rules and which serve a purpose that their participants value.7 Take the case of strategic coordination of economic behaviour in a competitive environment: Japanese firms lower the price of the cars they produce, thus driving German firms to do the same in order to remain competitive. This is a standard and ubiquitous case of economic interaction, yet nothing in the example, as such, suggests that Japa-nese and German firms regard, or ought to regard, each other as engaged in a rule-governed practice with a common purpose. Relational PIA scholars would argue that the impact of a dynamic such as this on people's lives is enough for egalitarian principles of justice to apply. PDA, instead, would contend that we first need to understand whether German and Japanese firms are actually engag-ing in a shared, rule-governed practice (e.g. through the rules of the WTO), whether the purpose of such a practice is to specify the terms of fair international economic competition, and whether the phenomenon described above goes against its purpose. If not, the fact that some agents are made worse off by some aspects of the practice is not per se an injustice, as long as the practice has a purpose that people can value8 and this purpose is endorsed and upheld by its participants. Second, most PIA theorists (of the relational variety) share with non-relational cosmopolitans, such as global luck egalitarians, the view that there is one conception of justice, strongly egalitarian in kind, and that the content of this conception does not depend on the nature of the practice that has to be regulated (e.g. Abrizadeh 2007). It is just that, for them, this conception only kicks in when a certain threshold of mutual interdependence is overcome, as opposed to what global luck egalitarians would argue (Schemmel 2007). PDA, instead, argues that the very content of a conception of justice depend on the nature and structure of the practice that it is going to regulate.

Now, arguments based on PIA in favour of global egalitarianism have often been criticized for failing to develop reasonably precise guidelines for reform of the situation of global inequality that they so ardently object to - both in terms of identifying which steps should be taken, and of figuring out which actors ought to be responsible for taking them. Whether this criticism has any force against theories that seek to deliver fundamental principles of justice is, however, uncertain, if it is pitched at the level of application: it seems that PIA theorists can simply acknowledge the point, and argue that the task of applying their prin-ciples is, of course, a difficult one that requires close collaboration with empiri-cal social science.

The objection we raise here is more fundamental: we argue that principles of global equality developed on the basis of PIA cannot even be evaluated as to their very soundness, because the means necessary to do so are lacking; PIA has

52 A. Banai et al. a methodological problem of justification, not merely a problem of application. Why this is the case becomes clear once principles of justice are understood as prescriptions that ought to guide our actions in a sufficiently clear way.9 On views based on PIA, institutional arrangements are wholly instrumental to the fulfilment of global egalitarian principles. So, on an understanding of principles of justice as action-guiding, the natural way of testing whether the principles that PIA views have developed are indeed sound would be the following: first, we take a set of differing institutional arrangements such as a world state, compre-hensive global governance (understood as a horizontal, decentralized web of global institutions), the state system, etc. - all of which should be, in principle, accessible and ascertain which of these would, as a whole, realize the princi-ples best. Then comes the test: we double-check whether we would indeed choose that arrangement over the others, or whether, based on our knowledge of how this arrangement would work, we become aware of cogent moral objections to it that, in turn, may lead us to question the very principles that it is supposed to implement, and to favour other principles; for example, principles that do not require a strong form of global interpersonal equality. Now, the problem is that, for all we know, such a test is not available: the knowledge about the compara-tive merits of different schemes of global institutions needed to carry it out does not exist.1 0 If this is so, then there is no adequate way of testing our allegiance to the principles developed on the basis of PIA. PDA, on the other hand, given that it takes existing practices as partly constitutive of, and not merely instrumental to, justice, does not require such unavailable tests, and is hence at a methodo-logical advantage.11

Let us elaborate on this objection in more detail: as is often noted, PIA's prin-ciples of global equality commit us to questioning, first and foremost, the exist-ence of states as we know them. In order to know whether we should subscribe to these principles, we would have to know how PIA principles would ultimately handle states: how would they reform them, or would they abolish them? It is therefore problematic that global egalitarians, when making proposals on how to implement reforms that will get us closer to the state of global equality that they advocate, often draw on states, either implicitly or explicitly, as the agents that ought to bring about - and, most importantly, sustain over time - these reforms (e.g. Caney 2005a: ch. 5). Kok-Chor Tan, for example, subscribes to global egal-itarianism, understood in terms of the global application of principles of liberal egalitarianism such as Rawls' two principles of justice, but argues that this need not translate into calling for a world state (Tan 2004: 6-7, 93ff.). 1 2 In support of this claim, he points out that reforms such as, for example, the introduction of a global "Tobin Tax" on financial transfers, or of the Global Resource Dividend proposed by Pogge, do not require such a world state. But both these proposals are very modest; their implementation would fall far short of reaching anything like global liberal equality, as their original proponents have always made clear.1 3 If, on the other hand, what Tan and others really wish to advocate is a robust form of global interpersonal equality, it seems clear that this will be incompatible with the preservation of state institutions in a strong way. For the

Global social justice 53 point of having states is precisely to have sub-global communities with mean-ingful discretionary powers as to how to organize themselves, their economy, and their social life - thus allowing the possibility of international interpersonal inequalities to arise. Functioning states (albeit not self-sufficient in any strong sense) have unique and significant power to establish and maintain stable prop-erty regimes and set the terms of social cooperation with regard to both macr-oeconomic strategies and the fundamental set-up of welfare states and their social policies.14 PIA views hence cannot plausibly ward off the objection that they are not able to individuate reform proposals and the actors responsible for them with sufficient precision by claiming that, as a matter of fact, some variant of a global order based on states would best - even in ideal theory - implement global egalitarian principles of justice. Given the probability just mentioned that preserving meaningful state agency will over time lead to, or maintain, large socio-economic inequalities between individuals living in different states,1 5 there are strong reasons to question why exactly a state system should constitute the best implementation of global egalitarianism. The question "compared to whatT continues to impose itself. Thus, the problem stands: how do cosmopolitan egali-tarians envisage global political and institutional transformation? Without knowing that, we do not know whether their principles are plausible or not. 1 6

Note that the problem is not that states might not be willing to bring about these reforms. This really is an issue that should not be raised against political theorists doing ideal theory; to the contrary, their ideal theories might help clarify that it is only lack of political will that stands in the way of morally desir-able reforms. The problem is instead that it is not clear how and why cosmopol-itan proposals for global institutional reform that draw on states' agency follow from their cosmopolitan fundamental principles. If the point of PIA is to be action-guiding, and if its prima-facie advantage over PDA is its potential to criti-cize the very existence of states, then why do so many cosmopolitans end up relying on fairly modest proposals which leave states with a large amount of dis-cretion? One would want to know how exactly this follows from PIA. If it is a piece of non-ideal theory, based on the assumption that abolishing states or reforming them radically is not going to happen under non-ideal circumstances, then how is this connected to the level of fundamental principle? What is, even broadly, the political ideal that we are partly sacrificing in subscribing to such non-ideal proposals? By which road have we come to conclude that it is exactly these proposals that follow from them, and not others?

This is particularly relevant as it is not clear that views based on PDA would reject most of the concrete (and, as already mentioned, relatively modest) pro-posals advanced by PIA scholars, whether at the ideal or at the non-ideal level. Given the undisputedly dire state of global affairs in terms of poverty, and the pervasive and systemic power inequalities between global actors, which often contribute to depriving many states of their very capacity to act qua states1 7 it seems that most plausible moral views can agree on reforms that would bring about more global equality both between states and between individuals across borders, compared to the present state of affairs. 1 8 Hence, the plausibility of such

54 A. Banai et al. proposals, if they do indeed follow from PIA views, does not necessarily support them over their rivals.

Another reason why global egalitarians might be reluctant to put forward spe-cific, bold and radical global political reforms is the fear of global despotism (Orru and Ronzoni forthcoming). Most global egalitarians, following Kant, profess to abhor a world state, because it would supposedly be tyrannical.19 The objection raised here against PIA views, however, is not of this sort. It is meth-odological, not moral. It applies equally to other possible radical proposals based on PIA, such as the proposal to transfer what used to be regarded as essential state functions to a system of horizontal global governance, understood as a decentralized global institutional web. The objection is quite simply that, while there is relatively reliable knowledge regarding the effects of having some supra-national institutions, such as the EU and the WTO-GATS, taking over some tra-ditional state functions, there is no reliable knowledge as to how a world state or a non-state-based global order would fulfil global egalitarian principles, nor of how, through which set of feasible reforms, such orders could, at least in prin-ciple, be accessible to us. 2 0 Short of such knowledge, we do not know whether or not we ought to subscribe to principles that seem to direct us towards such pro-posals. The point is not, therefore, that PIA views are too demanding or idealis-tic from a substantive point of view, but rather that they are methodologically wanting (if they aspire to be action-guiding, which we assume they do), in that they do not provide us with the tools necessary to assess the plausibility of the principles that they put forward.

PIA hence sets itself too ambitious a task: developing principles of justice without drawing on any facts about basic practices that exist in the global arena, and hence without any eye to how, and to which agents, cosmopolitan principles of justice would be applied, ultimately leaves the principles themselves without adequate support. In the absence of such information, agents lack good reasons to subscribe to them.

Some adherents of PIA do not actually claim that their principles of global egalitarianism are supposed to give comprehensive guidance as to what kind of international order we should aim at. They make the more modest "pluralist" claim that they have identified at least one value that should inform such an order, but that might have to be traded off against others, such as national self-determination (Caney 2001b: 129).21 In so doing, they repeat a move that is made by prominent luck egalitarians, who claim that equality, or egalitarian justice, is only a pro tanto value, not an overriding one, and has to be balanced with other values, such as liberty, or efficiency (Cohen 2008: 271; Temkin 2000: 129-30).

However, this "pluralist" position incurs the same problem, mutatis mutandis, and may even incur an additional one. Following what has been said above, in order to have good reasons to accept a moral principle, we need an idea of what the implications of this principle may turn out to be. We might have intuitions that support the principle, but if these intuitions are to be subject to at least some possible testing and revising, they need to be brought to engage with others sup-porting different considerations. Specifying the principle itself in more detail is

Global social justice 55 hence only of limited value; what is ultimately needed is an idea of how this principle would perform in trade-offs with others. And thinking in the abstract about the range of trade-offs that seem acceptable to us is not of much help either: we need to know which options for trade-offs exist in the social context to which the principle in question ought to be applied. In the case of global justice, the question is how the possibilities of such trade-offs would be struc-tured and constrained by different possible sets of international institutions; we are hence back to square one. The additional problem is that, insofar as such more modest proposals imply that justice is just one value among others, it is at odds with a prominent understanding of the nature of justice, according to which justice consists of rights and duties which do not regularly have to be balanced against other values, but override them, at least in most circumstances, even at significant cost.

Having thus assessed the methodological weaknesses of PIA as an approach to justice, we can now return to PDA. As opposed to PIA, PDA can deliver "guid-ance where guidance is needed" (Rawls 1999a: 18); the principles of justice it identifies serve as constraints on existing social practices that cannot normally be overridden by other considerations. Even ideal-theoretical variants of PDA have a much better chance than PIA views of coming up with practically relevant recom-mendations which can be sensibly evaluated, because they apply to (idealized ver-sions of) existing practices about which we have by definition enough knowledge to carry out this evaluation - otherwise the view in question could not have suc-ceeded at the stages of identification and interpretation of the practice.

The arguments advanced so far seem to support PDA in many respects. One possible objection, however, still stands, namely: is PDA not biased towards the status quo? This objection is sometimes formulated by critics as a moral objec-tion: they point out that, even if adherents of PDA should succeed in showing that standing in a certain institutionally mediated relationship, such as that of reciprocal cooperation in the provision of primary goods, indeed gives rise to egalitarian principles of justice, this only shows that such relations are sufficient for such principles to apply, not that they are also necessary. This seems to shift the burden of proof back to the proponents of PDA views, who now seemingly have to deliver a moral argument why non-participants should be denied equal-ity - which seems a bit of a nasty task. The objection of status quo bias seems to have particular force because, as pointed out above, most adherents of PDA have argued for statist conclusions, claiming that no practices yet exist beyond the state level that ought to be regulated by egalitarian principles.22 But the distinc-tion between PDA as a methodology and statism as its supposed normative con-clusion has to be taken seriously; there is no principled reason to think that PDA has to lead to exclusively statist conclusions. As the next section will argue, PDA has the resources to show how demanding principles of justice can be extended to international practices beyond the state, by calling for normatively desirable reforms of existing practices, their supplementation by additional cor-rective practices - or indeed the abolition of practices that stand in contradiction to other practices of greater normative significance.

56 A. Banai et al. Whilst PDA so far has been used to defend somewhat conservative conclu-

sions, we would like to suggest that it has the potential of serving emancipatory goals, in at least two ways. First, it can show how we live in a world of overlap-ping practices, each of which has its normative demands, and that not only the practice of the state gives rise to requirements of social justice. Overlapping practices, moreover, affect one another, and the principles of justice governing them must mirror this fact. Second, PDA need not be hostile to the establishment of new practices. In a world of overlapping and mutually influencing practices, the justice of some practices can be put in danger by externalities generated by other practices, or by other disturbing factors such as the behaviour of uncon-strained actors across different practices. In such cases, PDA will recommend the establishment of new practices in order to preserve or re-establish the justice of currently existing ones according to their own, internal and practice-dependent principles of justice. This would still be, however, a practice-dependent enter-prise, for (1) the new principles of justice would mirror the normative demand of restoring the justice of existing practices, rather than of implementing independ-ently arising global egalitarian principles; and (2) the existing practices would build on incrementalist and empirically informed accounts of what could help restore the stability and justice of already existing practices.

A world of overlapping practices As already noted, PDA has often been confused with straightforward statism and, consequently, charged with status quo bias. The charge, we believe, is to some extent unfair, and to the extent that it is not, it is a charge against specific interpretations put forward by specific variants of PDA, rather than against the methodology of PDA as such. Most advocates of PDA argue that, on the basis of the best interpretation of the state that can be provided, the state is the only social practice that triggers demands of social justice, that is, mutual relational demands among its participants (be these distributive or based on status, relative power, and/or social relationships). All other supra- and international practices do not trigger requirements of social justice in this sense,2 3 but only absolute demands that a decent minimum be guaranteed or interactional demands that certain standards be met in mutual interaction and behaviour. These are also the two demands which, as we have noted in the introduction, liberal statists would be prepared to accept. This, however, can be construed as an internal dispute within PDA, and it depends on substantive and empirical disagreements regard-ing the interpretive stage of the practice-dependent enterprise. On a different and more plausible reading, the moral map that PDA should draw is more complex. We live in a world of overlapping practices, and the specific rules of each prac-tice often have a major impact on the internal functioning of other practices. Whether a practice succeeds in serving its own purpose, and in being just accord-ing its own criteria, often depends on the impact of other practices. Therefore, the construction of principles of justice for all relevant practices should be undertaken with an eye to their mutual interdependence. For instance, the rules

Global social justice 57 of trade under the WTO should be devised in such a way as to take into account both the purpose of the relevant practice, namely global trade, and the impact of said practices on the internal functioning of other practices and institutions, such as the delivery of primary goods on the part of states according to sound princi-ples of domestic social justice. 2 4 If certain rules of global trade, such as TRIPS and escalating tariffs, prevent states from being capable of delivering primary goods to their citizens and residents in a just fashion, then they should be abol-ished or reformed for practice-dependent reasons. Note how this interplay between practices situated at different levels is something that we take for granted when we devise rules for sub-state rule-governed practices, such as con-tract law or the legal status and powers of clubs and churches; we devise rules that best serve the purpose of said practices, compatible with preserving the capacity of states to serve their purposes and respect their own duties of justice. If this is the case, however, several supra- and international rule-governed prac-tices trigger demands of justice over and above sufficientarian and interactional ones. For in order to pay attention, say, to the effects of the rules of the practice of global trade on the internal justice of societies, we have to pay attention to comparative levels of wealth above a bare minimum, and, most importantly, to the relative stands of different states in terms of bargaining power. Other supra-national practices, such as the EU, may trigger even more demanding require-ments, given that one of the shared aims of the practice is widely understood as being that of enhancing, or at least preserving, state capacity in an era of globali-zation. This is, to some extent, already recognized by the participants in the prac-tice. The structural and cohesion funds of the EU, for instance, have the explicit aim of reducing inequalities in wealth and opportunities between different regions of the EU. This is particularly significant because not only do the funds have a clear comparative and distributive aim, but they also do so at the regional level, thus clearly interfering with some intra-state affairs of their members. Thus, the marriage between PDA and statism is a contingent, and perhaps unhappy one.

Finally, there is a further, and highly important, way in which PDA can suc-cessfully respond to the charge of status-quo bias. The internal just functioning of a rule-governed social practice can, as we have seen, be influenced by that of other practices. Furthermore, especially at the global level, some powerful actors have the capacity of moving freely and escaping the control and regulation of any practices whilst, at the same time, significantly undermining the problem-solving capacity of some practices. The task of regulating corporations, making them accountable to the public interest and enforcing labour standards is, for instance, traditionally and almost uncontroversially understood as belonging to the state. As is widely known, however, global justice activists and scholars have powerfully argued that, under contemporary global circumstances, transnational corporations easily escape such regulation without being accountable to virtually anyone. They can buy a local company, copy a particularly successful techno-logy, and then transfer the production somewhere cheaper, thus creating shocks to the economy, destroying the social capital of specific regions, and eroding

58 A. Banai et al. long rooted systems of labour rights in established welfare states, or preventing them from emerging in the first place in economic more fragile states. They can create cheap and fragile jobs, safe in the knowledge that, if organized labour or even local institutions protest, they can move somewhere else where institutions and civil society are much weaker. We would like to suggest that PDA can identify what exactly raises problems of justice in such dynamics in a richer and more illuminating way than PIA approaches, be they relational or not. There is no supranational practice governing the behaviour of powerful economic non-state actors across borders - the ILO has virtually no enforcement powers, and the WTO governs rules of commerce among states only. However, these trends undermine the capacity of existing practices - most notably, states - to serve their purposes (the delivery of primary goods) in a practice-dependent, just way. 2 5 In similar cases, only the establishment of new regulatory practices can allow existing practices to preserve their justice (Ronzoni 2009). Even in such cases, however, proposals for global political reforms are likely to be developed in a different spirit with respect to views based on PIA. Reforms recommended by views based on PDA will be both more specific and more incrementalist. PDA views can allow for practice-independent relationships and dynamics, such as the one mentioned before, to be transformed into, or supplemented and regu-lated by, new rule-governed social practices. But such political transformation will necessarily rest on an incrementalist approach, based on the regulative idea of restoring the justice of currently threatened existing practices. Thus, attention will be given, not to unrestrained moral reasoning about the unfairness of the behaviour of powerful transnational non-state actors, but rather to the way in which their behaviour can be regulated in such a way as to no longer threaten the justice of the societies in which they operate. The content of the principles of justice that will regulate the new practices will depend on this, and therefore, unlike in the case of PIA, it will not be by default comprehensively egalitarian; this is what it means to say that the specific content of principles of justice will rather be determined in a practice-dependent fashion.

In conclusion, then, this chapter has argued that the methodological dispute regarding the role of existing social and political practices is at the bottom of current substantive disagreements in the global justice debate. We have gone on to argue that PDA is a promising method, not because it is more "realistic", but because it allows us to test its principles in a way that PIA does not. Finally, we have argued that PDA can be developed in such a way so as to make it both less status quo biased and more capable of adequately describing the complex sce-narios of the current global order. We have claimed that PDA need not be status quo biased for two reasons. First, PDA can and must describe the global order as a world of overlapping practices, whereby the rules governing each practice may play a major role in influencing the capacity of other practices to serve their pur-poses. Once this is acknowledged, space is allowed for principles of social justice governing practices other than the state. Secondly, PDA can recognize the necessity of establishing new practices, when current practices cannot pre-serve their justice unless transnational phenomena and externalities of other kind

Global social justice 59 are regulated, and thereby new practices are created. In both cases, significant space for global social justice is allowed, without necessarily jumping to the conclusion that this must automatically translate into a transferral of our prin-ciples of domestic justice to the global level.

Notes 1 By which we mean principles that set stringent limits to permissible socioeconomic

inequalities, even if not all of them require strict equality (distributive of otherwise). 2 This at least seemed to be the case in Rawls (1999a) (much less clearly so in Rawls

(1993)). 3 The following, slightly earlier, contributions to the global justice debate can arguably

be interpreted, retrospectively, as relying on the practice-dependent method (to differ-ing extents): Blake (2001) Nagel (2005); Cohen and Sabel (ed.) Julius (ed.) Freeman (2007: chs. 8 and 9).

4 For examples of PIA and PDA in the present volume, see the contributions of Simon Caney and Andrea Sangiovanni.

5 For the purpose of this paper we leave aside further reasons for favouring PDA over PIA - such as, for instance, meta-ethical scepticism towards the existence of inde-pendent and objective principles of justice over and above processes of mutual justifi-cation of principles on the part of agents within shared existing practices.

6 See also Rawls' (1999a: 25) suggestion that: "the correct regulative principle for any-thing depends on the nature of that thing".

7 Thus, we do not consider the distinction between PDA and PIA to be equivalent to that between "relational" and "non-relational" views on justice.

8 PDA excludes purposes that are clearly oppressive, such as slavery, see Sangiovanni (2008).

9 Some political philosophers consider the purpose of principles of justice not to be action-guiding, but rather to spell out our most fundamental commitments and ideals, even if these are unrealizable - or even all things considered undesirable - under human conditions, see for example Cohen (2008). We shall not discuss these issues any further here, since virtually all participants in the global justice debate explicitly claim their arguments to have fairly direct action-guiding implications.

10 A similar argument as to the impossibility to "imagine" the proposals of PIA on global justice issues in an action-guiding way (and therefore to submit them to our test) is made in Risse (2008). Our argument, however, yields less strongly statist implications than Risse's, as will become clear below.

11 We return to its supposedly complementary disadvantage - status quo bias - in the third section below.

12 Tan (2004: 94) gives a succinct formulation of PIA: "[W]e distinguish moral cosmo-politanism from institutional cosmopolitanism, and see that cosmopolitan global justice is premised on moral cosmopolitanism, but not on institutional cosmopolitan-ism" (emphases added).

13 The Tobin Tax actually does not have any distributive aims, egalitarian or otherwise. Though its revenues could be spent on development, its aims are (a) stabilization of global financial markets, and hence efficiency, and (b) increased national policy autonomy in macroeconomic and monetary matters, see Tobin (1996: xi, xiii).

14 They continue, grosso modo, to do so even where they are complemented by suprana-tional institutions, such as in the EU.

15 This is the case also because, as many have pointed out, large-scale redistribution from one state to another would seriously damage incentives for states to behave responsibly in socio-economic affairs; see only Rawls (1999b: 117).

16 Unless, that is, one takes their global egalitarianism to be a non-action-guiding ideal -

60 A. Banai et al. an option which, as mentioned above, we do not envisage here, as most egalitarian cos-mopolitans see themselves, either explicitly or implicitly, as setting up theories about "what we should do" rather than "what we should think"; see Cohen (2008:268).

17 In particular with respect to those states that are not, as a matter of fact, developing, but stagnating or worsening, up to the point of becoming failing states, such as a great part of sub-Saharan Africa.

18 This is the thrust of the more recent works of Pogge, which aim to develop proposals for reforms eradicating world poverty that can draw, in an ecumenical fashion, on dif-ferent moral views - such as cosmopolitan egalitarianism and libertarianism - for support, Pogge (2008a).

19 However, for an argument in favour of a world state see Cabrera (2004). 20 Of course, it would unfair to raise any possible obstacle to cosmopolitan reform as an

objection to PIA - proponents of PIA may idealize here to some extent, as proponents of PDA do when they identify and interpret social practices (see above). But PDA theorists accept some general feasibility conditions for conceptions of justice, such as that these be publicly knowable, generally incentive-compatible, possess mechanisms that deliver assurance of compliance of others, etc. The idea is that some such set of conditions would also need to be fulfilled by cosmopolitan reform proposals in order to be judged in principle feasible.

21 See Armstrong (2010), who takes the trade-off thesis as a starting point, but argues that global egalitarianism can be much more accommodating of national self-determination than currently thought (employing, however, a very broad definition of global egalitarianism, which includes even views that see states as the primary sub-jects of global justice). See also Armstrong (2009: 313).

22 For criticism, see Barry and Valentini (2009); Armstrong (2009: 310). 23 There are notable exceptions: see for instance James (2005a). 24 For a practice-dependent account of the demands of justice triggered by the WTO, see

Clara Brandi's contribution to this volume. 25 This case is powerfully made, with respect to health care in particular, by Gillian

Brock and Norman Daniels in their contributions to this volume.

Social Justice, Global Dynamics Theoretical and empirical perspectives

Edited by Ayelet Banai, Miriam Ronzoni and Christian Schemmel

R Routledge Taylor £> Francis Croup

LONDON AND NEW YORK

First published 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon 0X14 4RN Simultaneously published in the USA and Canada by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2011 Ayelet Banai, Miriam Ronzoni and Christian Schemmel for selection and editorial matter; individual contributors their contribution. The right of the editors to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Social justice, global dynamics : theoretical and empirical perspectives / edited by Ayelet Banai, Miriam Ronzoni, and Christian Schemmel. p. cm. - (Routledge research in IR theory) Includes bibliographical references and index. 1. International cooperation. 2. Social justice. 3. Distributive justice. I. Banai, Ayelet. II. Ronzoni, Miriam. III. Schemmel, Christian. JZ1318.S635 2011 327.1-dc22 2010041817 ISBN: 978-0-415-57569-0 (hbk) ISBN: 978-0-203-81929-6 (ebk) Typeset in Times by Wearset Ltd, Boldon, Tyne and Wear Printed and bound by TJI Digital, Padstow, Cornwall